Scale: The Universal Laws of Growth, Innovation, Sustainability, and the Pace of Life in Organisms, Cities, Economies, and Companies

Book Review · Complexity · Scaling Laws

The Arithmetic of Bigness

Geoffrey West's Scale claims that mice, elephants, Lisbon and General Motors are all running the same equation. The claim is bolder than it first sounds — and the ending is bleaker.

August 2026 · A standalone review, outside the trail series

Start with a fact that should be stranger than it is. Take a mouse and an elephant. The elephant is roughly two hundred thousand times heavier, and you might reasonably expect it to burn two hundred thousand times as much energy. It doesn't. It burns about ten thousand times as much. Per kilogram, the elephant is far cheaper to run than the mouse — and this holds, with unnerving regularity, across essentially every animal anyone has ever put on a scale.

Now the sequel. The elephant's heart beats about thirty times a minute; the mouse's beats six hundred. The elephant lives decades; the mouse lives a couple of years. Multiply rate by lifespan and both come out at roughly the same number of heartbeats — somewhere around a billion and a half. Size does not merely change how big an animal is. It sets the tempo at which the animal lives and dies.

Geoffrey West, a theoretical physicist who spent his career on quarks before turning to biology and later presiding over the Santa Fe Institute, spent thirty years asking why. Scale, published by Penguin Press in 2017, is his answer — and then, in its second half, his considerably more contentious extension of that answer to cities and to companies.

The book in one line

Big things are not small things made bigger. They obey power laws with specific, repeating exponents — and those exponents fall out of the geometry of the networks that keep the system supplied, whether those networks are blood vessels, water mains, or human relationships.

A physicist walks into biology

The intellectual ancestor here is Galileo, who noticed you cannot simply enlarge a man into a giant: volume grows as the cube of length while bone strength grows only as the square, so the giant collapses under himself. That is the crude version of scaling. West is after the subtle version — the cases where the exponent is not a clean 2/3 or 3/2 but something odd, and the same odd number keeps turning up.

The canonical example is Kleiber's law, named for the Swiss agricultural chemist who reported in 1932 that metabolic rate scales as body mass raised to roughly the three-quarter power.[1] Four decades of biology treated that 3/4 as an empirical curiosity. In 1997, West, working with the ecologists James Brown and Brian Enquist, published a model in Science that claimed to derive it.[2]

The derivation rests on three assumptions about the network that distributes resources through a body:

Assumption	What it means
Space-filling	The network has to reach every cell. No tissue can be left unplumbed.
Invariant terminal units	A capillary in a shrew is essentially the same size as a capillary in a blue whale. The last branch does not scale.
Optimisation	Evolution has minimised the energy wasted pushing blood through the system.

Impose those three constraints and you get a branching, approximately fractal network — and out of the mathematics falls the exponent 3/4, along with a whole family of quarter-powers: heart rate scaling as mass to the minus one-quarter, lifespan as mass to the plus one-quarter, growth rate, aorta radius, tree-trunk thickness. The invariance of lifetime heartbeats stops being a party trick and becomes an algebraic consequence.

M^3/4Metabolic rate versus body mass — Kleiber's law, across 27 orders of magnitudeWest / Brown / Enquist

~1.5bnHeartbeats in a mammalian lifetime — roughly invariant from shrew to whaleScale, ch. 3

~10 yrHalf-life of a publicly traded US company, largely independent of sectorDaepp et al., 2015

Why you stopped growing

The most elegant payoff in the biological half of the book is an explanation for something so ordinary nobody thinks to ask about it: why do you stop growing? A child grows for fifteen years and then simply doesn't, and no external instruction arrives to say so.

West's answer is arithmetic. The energy your network can supply scales sublinearly, as mass^3/4. The energy required merely to maintain the cells you already have scales linearly, as mass¹. A sublinear curve and a linear curve starting from the same origin must cross. At the crossing point, everything coming in is spent on staying alive and there is nothing left over to build with. Growth stops — not because a timer expired, but because the books stopped balancing. The same logic, run forward, gives a network-damage account of ageing and death.

Sublinear scaling is why growth ends. Superlinear scaling is why cities don't.

Cities are not big companies

Halfway through, the book changes register and the risk goes up. With Luís Bettencourt and colleagues at the Santa Fe Institute, West applied the same log-log machinery to urban data from the United States, Europe, China and Japan, and reported two distinct exponents living inside every city.[3]

What is being measured	Exponent	Consequence of doubling the population
Infrastructure — road surface, electrical cable, petrol stations	≈ 0.85 (sublinear)	You need only about 85% more. Bigger cities are per-capita cheaper and greener. Economies of scale, exactly as in biology.
Socioeconomic output — wages, GDP, patents, professionals	≈ 1.15 (superlinear)	You get about 115% more. Roughly 15% bonus per head, per doubling. Increasing returns.
Socioeconomic pathology — violent crime, contagious disease	≈ 1.15 (superlinear)	The same 15% bonus. The good and the bad ride the identical exponent.

The mechanism West proposes is that a city's defining network is not its pipes but its people — the density of social interaction, which grows faster than headcount. That single move explains why cities have increasing rather than diminishing returns, and it explains, uncomfortably, why you cannot order the innovation without the crime. They are two readings of one number.

The most useful idea in the book, and it's a footnote

Because the curve predicts what a city of a given size should produce, the interesting quantity is the residual — how far above or below its own line a city sits. That is a size-corrected performance measure, and it is far more honest than the per-capita league tables that fill newspapers. West reports that these residuals are strikingly persistent: cities that over-perform tend to keep over-performing for decades.

Companies die. Cities almost never do.

Then the third case, and the one most likely to be quoted at you in a meeting. Analysing tens of thousands of publicly traded American firms, West and colleagues found that companies scale sublinearly — like organisms, not like cities — and that their mortality follows an approximately exponential curve with a half-life of about ten years, largely regardless of what business they are in.[4]

His explanation is the growth argument again in a suit. A young company is mostly exploration: new products, new markets, income exceeding upkeep. As it matures, the share of its activity devoted to administration, maintenance and internal coordination rises toward the share devoted to innovating. The sublinear supply curve meets the linear cost curve, growth plateaus, and the firm becomes acquisition bait or a bankruptcy filing. Cities escape this because their exponent is above one; companies do not because theirs is below.

Drop a nuclear weapon on Hiroshima and the city is back. Take the best-run corporation of 1955 and check whether it still exists. Both observations are in the book, and the asymmetry is the point.

The treadmill at the end

The final chapter is where Scale stops being a natural-history book and starts being an argument about the century.

Sublinear growth, as we saw, produces a curve that flattens. Superlinear growth does something else entirely: the equations run away to infinity in finite time — a mathematical singularity. Reality obviously does not do this, so something must reset the curve before it arrives. Historically, that something has been a major innovation: agriculture, iron, steam, electricity, computing. Each reset buys another run at the wall.

The sting is that the clock is set by the system's own metabolism, and that metabolism is accelerating. So each cycle must arrive sooner than the last. Innovation is not merely required — it is required at an ever-shortening interval, indefinitely. West's sustainability worry is not fundamentally about any single resource running out. It is that the machine demands we sprint faster forever to keep from falling over.

Readers of this blog will recognise the shape of that argument. It is the same structural claim as the one in The Second Derivative: the alarming quantity is not the rate, it is the change in the rate. West gets there from urban economics; the climate record gets there from radiative physics. That two unrelated routes arrive at the same worry is either a coincidence or the whole story.

Where the ice gets thin

A book this confident deserves to be read with a pencil. Three reservations, in increasing order of seriousness.

Claim	How well it holds up
Metabolic rate scales as M3/4, derived from network geometry	Contested, not settled. Re-analyses of mammalian datasets have argued for exponents closer to 2/3 — the plain surface-area answer — or for exponents that drift systematically across taxa rather than sitting on one universal value.[5] The book presents the debate somewhat less evenly than a neutral referee would.
Urban socioeconomic output scales as N1.15	Sensitive to where you draw the city. Using administrative boundaries, metropolitan statistical areas or functional commuting zones can move the exponent materially, and several replication efforts find it drifting toward linear under alternative definitions.[6][7]
Organisms, cities and companies are instances of one underlying law	This is the book's rhetorical engine and its weakest joint. Similar exponents are evidence of similar mathematics, not proof of a shared mechanism. Power laws are notoriously easy to fit and notoriously hard to attribute to a unique generating process. That vasculature and human sociability are the same kind of network is asserted more often than it is demonstrated.

And a stylistic one

It is a long book. West says openly that he is painting a big picture, and the Los Alamos and Santa Fe anecdotes are warm and often illuminating — but the core argument would survive a third fewer pages without complaint. Readers who want the claims without the memoir can go straight to the papers in the references; they carry the empirical weight.

The verdict

Read it for the biology, which is superb and largely uncontroversial in its broad strokes. Read it for the urban chapters, which will permanently change how you read a statistic about a city — but read them alongside the critics. Read the final chapter as a provocation rather than a forecast; the mathematics of finite-time singularities is real, the extrapolation to civilisational timetables is a physicist's flourish.

What survives all the caveats is the habit of mind, and it is worth the four hundred pages on its own: before asking whether a number is big, ask what it should be for something of that size. Most comparisons we make — between countries, cities, firms, athletes, animals — are secretly comparisons of size wearing a disguise. West's contribution is to hand you the correction factor and let you see what is left over. That residual, not the raw number, is where the interesting question always was.

If you liked this, read next

Jane Jacobs for the qualitative case for cities that West quantifies; Michael Batty on urban modelling for the rigorous counterweight; Vaclav Smil on energy and growth for a colder look at whether the treadmill can keep accelerating at all.

References

1. West, G. (2017). Scale: The Universal Laws of Growth, Innovation, Sustainability, and the Pace of Life in Organisms, Cities, Economies, and Companies. Penguin Press, 496 pp. ISBN 978-1594205583.
2. Kleiber, M. (1932). “Body size and metabolism.” Hilgardia 6(11), 315–353.
3. West, G., Brown, J. & Enquist, B. (1997). “A General Model for the Origin of Allometric Scaling Laws in Biology.” Science 276(5309), 122–126. science.org
4. Bettencourt, L., Lobo, J., Helbing, D., Kühnert, C. & West, G. (2007). “Growth, innovation, scaling, and the pace of life in cities.” PNAS 104(17), 7301–7306. pnas.org
5. Daepp, M., Hamilton, M., West, G. & Bettencourt, L. (2015). “The mortality of companies.” Journal of the Royal Society Interface 12(106), 20150120. royalsocietypublishing.org
6. Dodds, P., Rothman, D. & Weitz, J. (2001). “Re-examination of the ‘3/4-law’ of metabolism.” Journal of Theoretical Biology 209(1), 9–27. pubmed.ncbi.nlm.nih.gov
7. Arcaute, E. et al. (2015). “Constructing cities, deconstructing scaling laws.” Journal of the Royal Society Interface 12(102), 20140745. royalsocietypublishing.org
8. Leitão, J., Miotto, J., Gerlach, M. & Altmann, E. (2016). “Is this scaling nonlinear?” Royal Society Open Science 3(7), 150649. royalsocietypublishing.org
9. Bettencourt, L. & West, G. (2010). “A unified theory of urban living.” Nature 467, 912–913. nature.com

Published August 2026 · A standalone book review, outside the History of Trail Running in Portugal series · Companion piece to The Second Derivative.

On method and tools

This review was written collaboratively with Claude Opus 5 (Anthropic): human specification and critical review, machine synthesis and drafting. The book's central claims are cited to the primary papers rather than to the book's own summaries, and the section on contested findings draws on the published critiques of both the metabolic model and the urban scaling exponents. Where the book asserts more than the literature settles — the universality of the 3/4 exponent, the robustness of 1.15, the leap from network physics to social systems — that gap is stated plainly rather than smoothed over.

A good book is not one you agree with. It is one that hands you a better question.

Authored by: Luis Matos Ferreira
Physicist & Developer